Online conversations have become more prevalent on public discussion platforms (e.g. Reddit). With growing controversial topics, it is desirable to summarize not only diverse arguments, but also their rationale and justification. Early studies on text summarization focus on capturing general salient information in source documents, overlooking the argumentative nature of online conversations. Recent research on conversation summarization although considers the argumentative relationship among sentences, fail to explicate deeper argument structure within sentences for summarization. In this paper, we propose a novel task of argument-aware quantitative summarization to reveal the claim-reason structure of arguments in conversations, with quantities measuring argument strength. We further propose ARQUSUMM, a novel framework to address the task. To reveal the underlying argument structure within sentences, ARQUSUMM leverages LLM few-shot learning grounded in the argumentation theory to identify propositions within sentences and their claim-reason relationships. For quantitative summarization, ARQUSUMM employs argument structure-aware clustering algorithms to aggregate arguments and quantify their support. Experiments show that ARQUSUMM outperforms existing conversation and quantitative summarization models and generate summaries representing argument structures that are more helpful to users, of high textual quality and quantification accuracy.

翻译：在线对话在公共讨论平台（如Reddit）上日益普遍。随着争议性话题的增长，不仅需要总结多样化的论点，还需阐明其依据与论证过程。早期的文本摘要研究侧重于捕捉源文档中的一般性显著信息，忽视了在线对话的论证性质。近期的对话摘要研究虽考虑了句子间的论证关系，但未能为摘要生成揭示句子内部更深层的论证结构。本文提出一种新颖的论点感知量化摘要任务，旨在揭示对话中论点的“主张-理由”结构，并通过量化指标衡量论证强度。我们进一步提出ARQUSUMM框架以解决该任务。为揭示句子内部的潜在论证结构，ARQUSUMM基于论证理论，利用大语言模型的小样本学习能力识别句子中的命题及其“主张-理由”关系。在量化摘要方面，ARQUSUMM采用论证结构感知的聚类算法对论点进行聚合并量化其支持度。实验表明，ARQUSUMM在现有对话摘要与量化摘要模型上表现更优，生成的摘要能更有效地呈现论证结构，具有更高的文本质量与量化准确性，对用户更具帮助性。